Thursday, March 15, 2012

Climategate and the construction of scientific facts

Marianne Ryghaug and Tomas Moe Skjølsvold of Department of Interdisciplinary Studies of Culture, Norwegian University of Science and Technology in Trondheim have studied the e-mail correspondence, which became publicly known in what is named "Climategate". They do not value what they read but ask how common the documented social processes are within science. Read yourself:

The global warming of climate science: Climategate and the construction of scientific facts


Abstract
This article analyzes 1073 emails that were hacked from the Climatic Research Unit (CRU) at the University of East Anglia in November, 2009. The incident was popularly dubbed “Climate Gate”, indicating that the emails reveal a scientific scandal. Here we analyze them differently. Rather than objecting to the exchanges based on some idea about proper scientific conduct, we see them as a rare glimpse into a situation where scientists collectively prepare for participation in heated controversy, with much focus on methodology. This allows us to study how scientists communicate informally about framing propositions of facts in the best possible way. Through the eyes of Science and Technology studies (STS) the emails provide an opportunity to study communication as part of science in the making across disciplines and laboratories. Analysed as “written conversation” the emails provide information about processes of consensus formation through ‘agonistic evaluations’ of other scientists work and persuasion of others to support ones own work. Also, the emails contain judgements about other groups and individual scientists. Consensus-forming appeared as a precarious activity. Controversies could be quite resilient in the course of this decade-long exchange, probably reflecting the complexity of the methodological challenges involved.

To read the full article, download it here.. It appeared in This paper appeared in International Studies in the Philosophy of Science. Vol. 24, issue 3, 2010, p. 287-307. DOI: 10.1080/02698595.2010.522411

54 comments:

Hector M. said...

The corpus of emails that appeared in the Climategate events (I and II) cannot be fully understood, not even as a "conversation", unless they are supplemented with other materials outside the corpus. They include most notably the writings of people outside the group of Climategate correspondents, which are all on mostly one side of the controversy. Interpreting the emails without reading such materials as appeared in the Climate Audit blog, or the account of the Hockey Stick affair as reported by Andrew Montford in his book The Hockey Stick Illusion, or the goings within the community of IPCC coordinators, authors and reviewers, or the papers that all these people were publishing (or trying to keep from being published) at the time, is in my opinion misleading and fruitless for an analysis of the scientific issues surrounding recent developments in climate science and its reflection on the IPCC reports.

These developments are indeed an enormously important field of analysis for science studies. But they should look at the ensemble of facts and not to a limited and partial corpus of documents.

@ReinerGrundmann said...

Hector

have you read the article? If so, what makes you so dismissive of their method? If not, do you think that such a general methodological caveat is warranted?

Your statement smacks of some hidden agenda. Surely researchers will be allowed to work with the material they can access. To demand more in this case would mean not to conduct any studies.

MikeR said...

Well, I read it. Pretty empty, I thought. "Sometimes scientists aren't sure their theories correspond to reality." Exciting things like that.

I agree with Hector, by the way. Steve McIntyre, for instance, has some interesting posts, where he shows a timeline, events from his side vs. emails from climategate. They helped me form a better picture of what the scientists were trying to do. I don't see why someone studying the event shouldn't use the broader world of information surrounding the event.

MostlyHarmless said...

Interesting, but Ryghaug and Skjølsvold fail in their analysis of the infamous "Mike's Nature trick" email. They note that Phil Jones said earlier "It is possible to add the instrumental series on from about 1980 (Mike [sort] of did this in his Nature article to say 1998 was the warmest of the millennium - and I did something similar in Rev. Geophys.)" - he's talking about a proxy reconstruction with real temperature data "added in", in other words replacing proxy data which didn't reflect reality. They think this was "an effort to establish more and stronger associations between data that does not obviously provide the same information". The proxy data clearly didn't "provide the same information" because it was replaced with temperature data. They don't appear to see anything wrong in that.

They then look at the infamous "Nature Trick" email: "I've just completed Mike's Nature trick of adding in the real temps to each series for the last 20 years (ie from 1981 onwards) [and] from 1961 for Keith's to hide the decline". They say "Considering the informal and oral tone, the mail probably just described that a particular method (Mike’s trick) had been used to establish associations between different types of data, an act prompted by lack of post 1980 proxies and the diminishing correlation between observed temperatures and certain tree-ring data". Indeed not - this was exactly the procedure previously outlined earlier by Jones, and the "adding on" was a euphemism for "replacing", hardly "establishing associations between different types of data".

We've seen that graph - have Ryghaug and Skjølsvold never seen it? The proxy series were truncated, and a separate temperature line added in, in a spaghetti-tangle of lines that took high magnification to see the "trick". The proxy series ended in 1980 and 1960, corresponding with the start of the temperature line from 1961, this is fine if it's made clear on the graph, or is prominent in the text of the paper (or both), but neither of these things was done. That's scientific fraud, purely and simple, and Ryghaug and Skjølsvold missed it because they didn't examine anything other than was in the emails, and gave the scientists the benefit of the doubt. If they were conducting an in-depth and unbiased analysis of the emails, then they also failed on this second count: giving them the benefit of the doubt surely wasn't in their brief. Of course they weren't performing an unbiased analysis. as they said in the abstract: "Rather than objecting to the exchanges based on some idea about proper scientific conduct, we see them as a rare glimpse into a situation where scientists collectively prepare for participation in heated controversy, with much focus on methodology".

That "methodology" included at best deception, at worst scientific fraud, and on several occasions too - if it had been discovered in a proper "non-pal" review, the paper(s) would have been rejected. The bias of Ryghaug and Skjølsvold is plain to see, right from the start - that abstract. Their paper is a whitewash.

mheimann said...

Amazing that a stack of hacked emails is now apparently enough to form the scientific basis for an analysis on how climate scientists tick. In the past, scholars of the philosophy science had to interview the pertinent actors or (if especially if already deceased) had to sift through lab books, notebooks, published and unpublished work or reminiscences from contemporary colleagues. Nowadays a few emails suffice to make a story that can be published. Sheds a sober light on the standards employed in the field of the history or the philosophy of science...

There's of course a deeper issue here: how will a historian of science in 20 years from now asses our motives, opinions and judgements? Will he analyze just a few emails or blog entries? I very much hope not...

Anonymous said...

Gotta love the hyenas showing up, repeating long debunked stories.

MikeR amd Mostly Harmless may want to read the evisceration of McIntyre's storytelling on a blog called deepclimate. You can start here:
http://deepclimate.org/2010/05/11/how-to-be-a-climate-auditor-part-1-pretty%c2%a0pictures/

And follow that up with this one:
http://deepclimate.org/2010/05/14/how-to-be-a-climate-science-auditor-part-2-the-forgotten-climategate-emails/

Bam

@ReinerGrundmann said...

mheiman

Social scientists (including historians) use various data sources. Studying the correspondence between actors (even if limited in its scope) is a well established method, as is every purely textual analysis. Why should it be otherwise in this case?

And would you make the same objection to a climate scientist who studies only one tiny part of the puzzle? In your logic this could not be right because climate change is so much bigger.

Hans von Storch said...

I thought the analysis by Ryghaug and Skjølsvold relevant and interesting. For those of our readers, who know it all, it may represent a provocation, and from the reactions so far, it becomes clear that some are unable to look at the concept of analysis and its results in an open manner.

The two authors looked at the e-mails as data, and were trying to reconstruct what has happened. By leaving out the interpretations of others, of partisans of both types.

A key finding is: this is mostly "normal" (my quotation marks) business among scientist in pursuing their agenda of making the own results accepted as "truth". Thus, Ryghaug and Skjølsvold were surprised about the public response about the conduct, because that would be mostly normal in scientific circles.

I must say that my personal experiences do not tell me that such behavior would be normal in my scientific milieu. But, I am overseeing only part of the milieu, and I have to admit that there are these groups, who are mainly driven by the need to get funding, run the committees, and who are not always known for critical and innovative scientific analysis. Thus, I tend to believe Marianne and her coworker that they may be right, that the practice is indeed rather normal, and that I am living in an island of naivety. (But I am glad to do so.)

Now, if true, an interesting conclusion could be that this practice may not be uncommon in science but mostly unknown in the public (otherwise there would not have been the outrage, which we could observe). That the public believes that science would be Mertonian (read: Stehr, N. 1978: The norms of science revisited: social and cognitive norms. Sociological Inquiry 48:172)- which would mean that the public is funding a different activity than what it believes to do. Thus, the next step is not to decry Ryghaug and Skjølsvold, but to ask - which function, role, organization, quality control etc. does society expect from science, and how does society makes sure that science is acting according to these politically legitimated concepts of the public?

Werner Krauss said...

Some random thoughts concerning the second part of the article's title, "the construction of facts".

1) the article is about how a group of scientists "communicate(s) informally about how to frame propositions of facts in the best possible way". Facts are not to be discovered 'out there'; instead they are made / constructed / the result of activities as described in this article. This is a main message for both sides of the climategate dispute: there is no "truth" out there which can be discovered and which finally will decide the conflict. There are only good or bad constructions of facts about climate.

2) The "scandal" is not that we learn how a few scientists "try to keep von Storch out" (my main climategate email quote) and the like. Outsiders often think that universities are different form other workplaces. Quite the contrary is true. Science is a closed and hierarchical system and thus the stage for permanent intrigues, bullying, harassment etc., without any outside control. The only guarantee we have is that most of the "facts" are so often tested that in the end they are "black boxed" and indeed "true". Or, to phrase it in other terms, they resemble reality so much that we can take them for real. It works pretty well, but different from what many people might think.

3) there is one special thing about climategate which indeed is a matter of concern: we don't know whether the thing with carbon based anthropogenic climate change is made-up or not. I mean, of course it is "made" in the sense of constructed; one should say: we don't know whether this is construction which will hold the test of time or not. And currently (and I am afraid for a long time), there is no chance to decide ...

4) That's why it is really interesting to learn how "propositions of facts" about climate come into being. As Stephen Schneider once said, it is up to him how scary he represents his findings. He is right insofar as facts are always represented, they cannot speak for themselves. There is no neutral way of presenting them. The only real lie is to pretend that the facts you constructed speak for themselves. The don't. That's why the IPCC correctly adds "most likely" etc.

5) In the end, WE decide what kind of climate trouble we are really in. This, indeed, is frightening. The experts do the best they can - this article gives a small insight into a part of their activity - but in the end, we all have to be experts. As Buckminster Fuller once said: On spaceship earth, we are all crew, there are no passengers.

@ReinerGrundmann said...

Hans
"Thus, I tend to believe Marianne and her coworker that they may be right, that the practice is indeed rather normal, and that I am living in an island of naivety."

The problem is that they do not have the evidence for such a claim. And I doubt anyone has. My conclusion would be that they use this as an argument to condone the practices seen in the emails.

I have two forthcoming articles on 'climategate', one of them will be published by WIRES Climate Change and has a section dedicated to Ryghaug and Skjølsvold. I will post them asap.

Werner Krauss said...

@ Hans #9

Oh, obviously we wrote and posted simultaneously! I really appreciate that you acknowledge this effort from part of science studies.

Just a short answer: you ask "how does society make sure that science is acting according to these politically legitimated concepts of the public"?

I do not fully agree that science is only a "service provider" for society. The public also funds art, for example, but nobody would suggest that artists have to fulfill the wishes of the audience. At least to a good deal, science is like art, too - exploring and representing the many past, present and future realities we inhabit, to put it poetically.

And we should keep in mind that the "public" is also only a construct and does not exist as such. The public is the result of sociological surveys, for example, which are, well, sociological surveys. It works, because the nation, the state, the administration, taxes, universities, schools work. There is a reality we can agree upon. But we cannot reduce life to this one reality. Politics know this, for example.

That's why I would argue against the exclusive "service mentality" of science. Science would be helplessly caught in the cage of its own governmental reality construction - eherm, sorry...sounds wild, but science has to construct reality, which is a creative (and not a service) process.

Climate is not only scientific, that's what maybe science has to learn. Climate is life, too. And we should not underestimated "the public", they don't expect too much from science. Already as children we learn the mocking song: "Die Wissenschaft hat festgestellt, dass Marmelade Fett enthält..."

Finally, I think that blogs are maybe a first step towards new forms of communication, where scientists have to act and argue like expert citizens - citizens, like all the others. Talk among crew members of spaceship earth, with different talents. Democratizing science does not mean reducing it to a mere service institution.

Hans von Storch said...

Werner, I did not say, and did not want to, that science would only be a service: that is indeed left to society, which function (for instance as service) science should have.

Totally different: For some reason, the stats counter is not counting this thread.

Dennis Bray said...

@Werner

Re: "And we should keep in mind that the "public" is also only a construct and does not exist as such"

What if we call them people instead of public? A construct, as you use the term is a theoretical creation based on observations which cannot be observed either directly or indirectly. It is ultimately a term we have agreed upon to use in communication. So, public, in this context seems to work - there is much agreement as to what it means. As an operational definition it is of course open to refinement - and this seems to be the weakenss of much of the cultural/post modern literature - too many concepts, no speciification. However, where the real confusion presents in itself in our corner, is the failure to distinguish between climate science and climate scientists. Science can be none other than neutral, it is a mere concept. Scientists, on the otherhand, give shape to science.

Werner Krauss said...

@ Dennis #13

Sure, "the public" is a well tested construct. It serves well for questions of the organization of a modern nation state, for example. You can count, observe, measure, survey, ask opinions of "the public". Indoing so, you turn them into "governable" subjects. But on the other hand, when political decisions or arguments are made "in the name of the public", we should be careful. It's like "nature" a bed fellow which too easily serves to hide your own interests.

(I always liked Steve Job's saying that people don't need to know what they want; he knows and sells it to them).

I see one of the successes of "postmodernism" (whatever that is exactely) that it indeed pluralizes our understanding of the public and makes clear that this construct is a) always related to power and b) is like a crystal ball - it somehow always reflects what you project into it. For example, a sociological survey delivers a different result than an ethnographic field study with the same population. This is no problem at all; it's just like that.
Plurality is not a danger; instead, it opens up possibilities.

You think science exists independently of scientists? I am not sure: isn't climate science what climate scientists do? If nobody gives shape to science, it doesn't exists anymore (or sleeps somewhere in a library and when discovered again, it is outdated). Or am I wrong here?

Werner Krauss said...

@ add to my previous comment:

to b) I have to correct myself; it does not "reflect what you projected into it" - that's too simple. For example the surveys that Reiner shows in the other post bring new results which were not anticipated. The rest is fine. Thanks.

eduardo said...

Specially Werner will enjoy this video

Werner Krauss said...

Thanks, Eduardo, I actually do enjoy both the article and the videos. It reminds of the origins of the constructivist epistemology in biology, by scientists such as Maturana or Varela.
Nice excursion into the theory of social constructivism in the klimazwiebel seminar -:)

Anonymous said...

Dear Hans, can I take your comments as claiming Phil Jones, Mike Mann, Keith Briffa and a few others featured in the climategate e-mails are just "driven by the need to get funding, run the committees, and who are not always known for critical and innovative scientific analysis"?

If not, you may need to clarify this, because this is certainly how it reads.

Bam

MikeR said...

@Anonymous #6. Aside from calling us names, what I see from these posts is that you find them very convincing ("evisceration"). I personally think this kind of thing is very valuable. McIntyre posts his version, deepclimate attempts a rebuttal, McIntyre has the chance to add more details, etc. The rest of us get to see more of the story and can make up our own minds. I wouldn't be at all surprised if McIntyre is right some of the time and Mann's group some of the time (though I can't tell you which way these examples go without tracking them down to the end).

On the other hand, one can instead just "pick a side", read their wonderful posts that make you feel good because they agree with you, read the other side's evil and irritating posts rarely or only as opposition research, and always of course conclude that one's own side is right every time.

MikeR said...

I have to say, Anonymous, on rereading the first of your two posts from deepclimate, that my mind boggles on your considering it an "evisceration". It seems to be focused on a point so trivial that most of us would consider it completely irrelevant (Was the green curve in the graph slightly under or over the other curves, at a level of resolution not visible to the naked eye? No, really), while leaving McIntyre's main points untouched.

I'd also add that one thing I appreciate about climateaudit is that he documents things. It is easy to check his work. If he quotes an email, the link to the email will be there. The same with his code, it's always right there in full. Your deepclimate links to his posts are pretty annoying, often to their own earlier posts (!), or sometimes to a google page.

Anonymous said...

MikeR, contrary to your claims, the issues pointed out on Deepclimate ARE McIntyre's main points in his climateaudit blogpost.

Also see how Mostly Harmless has picked up on what they call "the dog whistle" in that story of the spaghetti graph and supposed hiding of inconvenient data.

On to some fun trivia: I have on numerous occasions defended McIntyre's work in the early days of climateaudit (not on the Internet, there I only "lurked" at the time). Until I started to detect a growing tendency for narratives. Narratives have this tendency to look good, but on taking a second look contain inconsistencies, questionable claims, and frequently rather strenuous connections. There were more and more in McIntyre's blogposts, and that's where it stopped for me.

More fun trivia: I detected the narrative and its problems in McIntyre's post on the spaghetti graph before Deepclimate had blogged about it. I stumbled on him /her by accident, and saw someone else had dug even deeper, but in essence found the same. That made me feel good, yes. At least some people were no longer taking the McIntyrian narratives on face value...

Bam

MikeR said...

Well, I can't exactly comment on the "main point" of McIntyre's post, since, as I said, I found it very different to tell what post they were referring to. I eventually found a link they had to a pdf of some presentation he'd made, which wasn't much help. I distinctly recall that he had some very clear posts that went through the whole story (and I wish that deepclimate had linked to them), and I certainly had not had the impression that that was the major point. Maybe deepclimate was responding to some very trivial post? I would rather think that a graph like that is extremely misleading and extremely bad practice - either way. It "hides the decline".

Ulf T said...

Already on page 4, I start feeling bothered about the theme of the article:

Thus, controversy about scientific facts tend to become controversy about method, and the phenomenon of experimenter’s regress means that it is important not only to do good experiments but to be perceived to have proper methods and to be skilled in applying them. In turn, this means that it is vital to present one’s methods in a persuasive way.

As Hans observed, this may well be the position of "normal" (my quotation marks) business among scientist in pursuing their agenda of making the own results accepted as "truth", bit at the same time, these scientists often want to appeal to the authority of the scientific method and the consensus behind their findings as representative of unassailable truth.

Indeed the scientific method attempts to guide research towards 'truth', but then it is not about being perceived as having proper methods, but to openly and honestly describing your method and findings, and seek the help of your peers to determine if the results can be verified and replicated.

That this leads to 'controversy about method' is of course not strange; a vital part of the scientific method is to ascertain whether the method is sound, and that scrutiny should be brutally honest. But devising a method that unravels only after intense scrutiny is not a defeat - it's a very gallant attempt, as long as the difficulties in exposing the flaws were not a result of deception or interference on behalf of the original author.

In my opinion, what is being described Ryghaug and Skjølsvold is politics, not science. Science and politics are in conflict often enough that one should decide what to be and label oneself honestly.

Politicians should understand the scientific method, but they are not required to practise it. Scientists should understand (and are subject to) the rules of politics, but their first obligation is to the scientific method. When you put politics before science, you are no longer a scientist.

I do not call myself a scientist, but rather a subject matter expert. The above also holds for experts. I have long since learned that politicians have many uses, but have come to disregard 'experts' who misrepresent their own best understanding in order to gain political favor. Once you do, you no longer deserve the title, nor do you necessarily serve the politicians and decision makers who rely on your advice.

This is a difficult standard to live by, but there are many honest ways to earn a living (e.g. politics) where the ethical bar is ... erm, different.

Ulf T said...

...at least in my browser, the attempt to frame my quote of Hans von Storch in italics failed miserably. Apologies. Hans's words were:

« A key finding is: this is mostly "normal" (my quotation marks) business among scientist in pursuing their agenda of making the own results accepted as "truth". Thus, Ryghaug and Skjølsvold were surprised about the public response about the conduct, because that would be mostly normal in scientific circles. »

I agree that it does appear to be mostly normal, but argue that it shouldn't be.

Sven Türpe said...

@Ulf T:

I find the idea flawed that scientists would have a moral obligation to stick to "the scientific method". First, there is no such thing as a scientific method. Second, science as a system does not depend on the integrity of the individual scientist; science works as a system that emphasizes useful results and forgets about the less useful ones. The important mechanisms exist in the system, not in the individual scientist, and the system evolves over time. This, too, has science in common with politics: we don't need perfect politians, we only need a system that, statistically, promotes the good and demotes the bad.

Hans von Storch said...

Bam/18 - My line "that there are these groups, who are mainly driven by the need to get funding, run the committees, and who are not always known for critical and innovative scientific analysis. " had to do with the question if a behavior as documented in the ClimateGate mails would be "normal". It did not refer to the colleagues at CRU or wherever - this case was rather special, but the performance may possibly be "normal" in a broader sense of university cultures (as characterized in my line).

Anonymous said...

Hans, thank you for the clarification, although it actually leaves me more confused.

Bam

MostlyHarmless said...

"Anonymous" thinks I got my information from a climateaudit post. He magically knows what others have read and what their (unstated) opinions are. I didn't refer to anything outside the Ryghaug Skjølsvold paper, apart from the "spaghetti" graph, which I obtained myself, some time ago, and I have all the emails, downloaded the day after they were available on the 'net. I've read hundreds of them at least, and searched through the text many times. I make my own mind up on such matters as this, and do my own research - unusual in this day and age I know.

I'll summarise my point - "real temperatures", as mentioned several times in the emails, have no place on a proxy reconstruction, unless provided for comparison, and clearly identified on the chart, in its caption, or prominently in the text. This was not done. My dictionary defines "trick" as "An act or procedure intended to achieve an end by deceptive or fraudulent means". I can find no other interpretation to put on the word in this context.

There are many discussions in the emails about using different proxies, leaving out less convincing ones, and so on. That's all valid, though they might be accused of "cherry picking". Some scientists do this as a matter of course, and many of those state explicitly that they've done it, and provide justification. Full marks to them. The palaeo team knew that the divergent proxy series would open them to criticism that if the series were unreliable in the recent past, they might well be unreliable earlier also. They discussed this in the emails, which led eventually to Phil Jones suggestion of using "Mike's Nature trick".

Ryghaug & Skjølsvold clearly think that the "trick" was simply a method of improving the credibility of proxy reconstructions, where modern divergence reduced the credibility. Not so - the divergent tails were removed, to the chart didn't represent a proxy reconstruction at all, but literally a "cut and paste" fabrication. There was nothing whatever in the papers to identify this "trick", so the charts didn't represent what was claimed for them. That's scientific fraud.

DR_UK said...

Reading Ryghaug and Skjølsvold with, I hope, an open mind, I feel that their analysis could have benefited from understanding the context more widely.

For instance, they wrote in their Methods section “We have analysed the 1,073 e-mails” and did not mention (did they know?) that this was a subset – how it was selected has been a topic of debate. This was stated in the first release (though of course could not be checked) and confirmed by Phil Jones in his evidence to the House of Commons the Science and Technology Committee.

“You’ve only seen a tenth of 1 percent of my e-mails in this group.”

http://www.nytimes.com/cwire/2010/03/02/02climatewire-climategate-scientist-admits-awful-e-mails-b-66224.html

Of course it was also confirmed by the subsequent release of more emails in 2011, but that was after the publication of this paper.

Ryghaug and Skjølsvold should therefore be cautious about drawing quantitative conclusions (e.g. ‘many - nearly half - of the emails’) without recognising that this relates to an unknown subset of the whole email correspondence. It highlights the importance of understanding the context as broadly as possible. That context, as other commenters have noted, is available to the interested researcher.

Ulf T said...

@Sven Türpe,

I'm not sure what it was that you found flawed (not saying there can't be a flaw in my reasoning). Do you not agree that scientists have an obligation to honestly represent their work, rather than mould the results to increase their own financial payoff or fame?

I agree that the 'scientific method' is a very loose concept and that the system itself should be robust against bad science. This doesn't take away the individual responsibility of the scientist not to try to game the system.

I was once asked to help shape a software architect course for a major corporation. I was asked what I thought the most important characteristic of a software architect was. The one that came to me was 'courage'. You'll constantly be pressured to cut corners, due to time or cost pressures, and will have to stand your ground against people with more clout than you, but less knowledge about the working principles of the system your building. You may find yourself arguing with a whole lineup of superiors, having only your gut to tell you that what they're proposing is wrong and will come back to haunt you later.

The career software architects will stuff their own intuition and good judgement and make the career move (placate their bosses and move on before the troubles begin). The great software architects will dig in and find a way to explain what their superiors need to know.

The people best equipped to tell companies how to know a great software architect from a bad one are the software architects themselves.

In the same manner, scientists should be religious about what makes a great scientist. They are the system.

Ulf T said...

Related to this, perhaps, ClimateAudit just linked to another article studying the ClimateGate emails:

Ad hominem arguments in the service of boundary work among climate scientists (Souder & Qureshi, JCOM 11)

Still reading...

Sven Türpe said...

@Ulf T:

Exactly, I do not agree that scientists have such an obligation, or at least none that would specifically pertain to their role. They may have a general obligation to limit the amount of trouble they cause to others in their social network. This obligation surely suggests they should refrain from knowlingly misleading others. However, this obligation is in no way specific to science. We would expect the same of a software architect working in a team, of a stranger on the street whom we ask for directions, or indeed of anybody else.

I believe honesty is more of a courtesy to fellow scientists than a requirement for science to function. What happens if somebody knowingly pollutes the scriptures of science with dishonest representations? Essentialy the same thing that happens when an error goes unnoticed or a published hypothesis turns out wrong: a few people will have spent time and resources on something that didn't work out -- and they may still have learned things in the process.

Furthermore I'd like to stress that the very concept of honesty makes sense only for a subset of science. One may be dishonest about experiments and data and maybe about interpretations of data. But how could one be dishonest about such things as hypotheses or theories? What would a dishonest theory of relativity look like?

We must never forget that science is a continuous process. What gets published is not the results of science, what gets published is only what scientists communicate to other scientists. History determines the results of science: that which prevails and gets built on. Therefore there is nothing wrong with, for instance, the computer science approach of just making things up all day, some to be forgotten, some to be fixed later, and some to end up creating value at multi-billion dollar businesses like Google or Microsoft, which kind of demonstrates their merit. If we learn something in the process, it's science.

Anonymous said...

Mostly Harmless tells us he found out all of this himself. In that case, he clearly missed the clear labels on the spaghetti graph. That the "real temperatures" do not belong on a proxy graph is complete nonsense. Part of those "real temperatures" are used for the calibration of the proxies, and there is no reason to exclude them.

Moreover, you may want to check your dictionaries again.
You can start with the Merriam-Wesbter, which includes in its definitions of "trick":
"a quick or artiful way of getting a result"
Or the Oxford Dictionary, which includes
"a clever or particular way of doing something"
Also the Cambridge dictionary does this.
In math you have something called the "replica trick", which is a clever way of simplifying otherwise complex calculations.

To add injury to insult, Mostly Harmless clearly has never read the papers referencing Briffa's reconstruction. It is, in fact, a paper that discusses the divergence problem at length, including an explanation why removing the data after 1960 is taking place.

Bam

Hans von Storch said...

Bam/33 - you are making it too easy for you. You have to distinguish between what is written in technical papers by some, and the message which is received by the public through mediators, vested interests, blogs. When the scientists allow for misunderstanding - and I would guess this can be said in case of the spliced "trick"-curve - then they are responsible for confusing the public. Mann's "Nature-trick" has done a lot of harm, and is scientifically unacceptable.
Technically these scientists are right, but "Mostly Harmless" is also right.

Your pointing to the divergence-paper of Briffa is also fine, but as you seem to have read everything, you are certainly aware of an assertion given at the NRC-council (on the hockeystick) about this issue, namely that this issue is really a reason for concern, in particular if the link between temperature and tree ring characteristics may have been broken down before.

You will be able to tell us in one of your next comments here. The trick is to point to all relevant pieces of evidence, not only those you know.

The assertion "Mostly Harmless clearly has never read the papers referencing Briffa's reconstruction." certainly applies to all participants of this debate, when we replace "referencing Briffa's reconstruction" by "relevant here".
Please try to restrain your arrogance a bit.

ingno said...

A very interesting discussion!

In my mind it is also very revealing, not som much about how science work but about how STS is conducting their studies.

For example, the frase "This allows us to study how scientists communicate informally about framing propositions of facts in the best possible way."

How do they know that these letters reveals anything about regular scientific communication? I would rather agree with Hans von Storch that it is quite extra-ordinary.

The problem with STS-studies, as I see it is that they never seem to bother about the scientific problems as such; papers, conferences, books etc. Also in other disciplines. That would have helped them to make a more balanced judgement of whether the climategate letters are a good object "to study how scientists communicate".

Ingemar Nordin

Werner Krauss said...

@Ingemar Nordin

There is something about your remark that made me pause:

you want STS to bother more about "scientific problems as such" and demand a "more balanced judgment".

On the other hand, you take one article to discredit a whole discipline: you find the article "revealing (...) about how STS is conducting their studies".

I think the article deserves to be discussed in detail; to make it just part of a group and discredit the group as such is not really fair and "balanced". And it does not make go away the fact that reaching consensus via emails is a "precarious activity", as the authors state. It is precarious even among some of the most respectable and influential climate scientists, as the huge sample of emails shows. This is indeed "extra-ordinary", but maybe the "extra-ordinary" has been (and maybe still is) the new "regular" in the heated climate debate?

This does not mean that your critique is not valid; maybe some of the conclusions are indeed too general; but this is not a reason to dismiss the whole article with an authoritative gesture ("typical STS").

MikeR said...

http://climateaudit.org/2012/03/19/vignettes-before-mm2003/
More ad hominem!
I don't understand how people can read these things and defend the Mann side. Unless McIntyre has his facts wrong, and as usual they are carefully documented. And McIntyre has dozens of posts just like this one.
Reading them tends to upset me, with a feeling of helpless frustration: Just make your @#!?% data and methods public! What is the sense in this decade-long struggle against making enough information available so that others can actually check your work?
The only way I can understand it is a group of fairly normal scientists, suddenly thrust into world prominence - but who have pretty poor data-handling techniques, and pretty poor statistical techniques. And they can't deal with that being exposed, and they'll never admit it.

Shibui said...

MikeR has it exactly right.

ingno said...

Werner Krauss,

Yes, my conclusions about STS may be too general; it is based on my own unsystematic experiences and reading, and not only on this particular paper by Ryghaug & Skjölsvold.

However, I find their approach quite disturbing in that they take the Climategate letters as typical for how social processes and informal communication among scientists is conducted. They do this when it should be quite obvious to every external observer that the situation is extra-odinary in a larger politizised context: they are obviously considering stratagems for presenting their results to journalists and policy makers, not to other scientists.

And they claim that they use the STS paradigm: "Through the eyes of Science and Technology studies (STS) the emails provide an opportunity to study communication as part of science in the making across disciplines and laboratories."

But they are sadly mistaken if they believe that they "through the eyes of ... (STS)" are able to study "communication as parts of science in the making".

Ingemar

Anonymous said...

Hans, tell me, how can I, or anyone for that matter, point to pieces of evidence that I apparently do not know?

I am aware at the concern aimed at the bristlecone pines and foxpines (IIRC), but those are different from the Briffa proxy. Moreover, and again IIRC, the concern was not whether that link had been broken down before, but rather that most people did not use these proxies after 1850 or so, while Mann did. Funny, that discussion, considering that others then complained about the Briffa prxy being cut off after 1960!

Finally, there have been plenty of papers, both before and after the NRC report, that have investigated the divergence problem and come to the conclusion that it is an issue most likely only relevant to recent times. But maybe these are just the pieces of evidence I know...

Bam

P.S.: apparently, calling something "fraud, purely and simple" is not arrogant...(or at least not arrogant enough to get a response). Good to know. Goodbye.

Werner Krauss said...

Ingemar,

I still don't fully get your point. Here my short take on the article, what I like about it and what I miss. I am curious to learn whether it meets your critique or not.

The authors explicitly state that the emails provide "a very rare glimpse into a situation where a group of scientists collectively prepare for participation in a heated controversy" (2). And according to Knorr-Cetina, "controversies are windows of opportunity to study facts in the making". Thus, I consider their approach to the climategate emails as perfectly legitimate.

And they also consider the wider context such as how scientists prepare outside communication; how they are expected to speak "with one voice"; how they write papers; how they discuss internally; how they make facts credible and finally reach consensus. To make a long story short, I think this is good "old school STS". We get an insight how science works, how facts are constructed and consenus is reached.

Obviously, the authors consider science and consensus building in science as a pretty tough business; they seem not really surprised about the roughness of the email "gossip" and its tone: "Some of the emails may contain statements that, when cited out of context may look suspicious. However, (...) this appears as scientific business as usual." Well, that's how it is.

But there is something missing. Of course, it's interesting for some people to learn how scientists construct facts, and for some it may come as a surprise that educated professors are not too different from other people in leading positions. But this is not special for climate science.

What I miss is that they are not really interested in climate change itself. They pretty well show the methodological problems in climate science consisting of so many different disciplines. But the scandal is contagious; it keeps the authors' attention focused on it.

We don't learn what climate change really means for scientific routine; what challenges it poses and how science has to change accordingly; instead, the paper just seems to suggest that we have to do with business as usual - which in my opinion is not the case. Climate science in the anthropocene is no longer business as usual, because the challenge is new, and it's huge.

Werner Krauss said...

add:
and I forgot to add that there are many STS related studies which deal with climate science issues beyond those discussed by Marianne Ryghaug and Tomas Moe Skjølsvold, from Sheila Jassanoff to Mike Hulme, from Jerry Ravetz to Roger Pielke jr. and many others in-between.

Hans von Storch said...

Werner/41. "Well, that's how it is." - This is what the authors suggest as hypothesis, but is there more evidence from other studies - results showing similar results from other, possibly less spectacular cases? If I compare this hypothesis with my scientific environment, I would say: No, it is not "business as usual".

"Climate science in the anthropocene is no longer business as usual, because the challenge is new, and it's huge." - true, the challenge is new - for science and society - but is it "huge" for science", or perceived as "huge" by society?

The formulation "Climate science in the anthropocene" is likely not what you meant, but something like "Climate science dealing with the anthropocene"?

Now, how common was such extraordinary science/challenge, real or perceived, in the history of science: HIV, racism/human anthropology, Lysenkoism ...?

I find your suggestion that climate change would have an impact of how climate science is done, intriguing. This impact, has hardly been studied so far. But which "climate change" do you mean? - I would presume a culturally constructed CC (quadruple C).

Werner Krauss said...

@MikeR #37

"And they can't deal with that being exposed, and they'll never admit it."

This sounds too much like a lover's lament. She will never come back, she will never apologize, even though you have her diaries and proof that she cheated on you.

There is something strange about this. People involved stick to their narratives, climategate becomes part of their identity, the center of their world - not really sustainable, because everybody else will get tired of this story which knows no resolution.

Many folks said that Jerry Ravetz failed with his "reconciliation in the climate debate" project. I think he didn't. It 's only that some people didn't get the idea what "reconciliation" means.

Climategate is one step in the long history of the climate debate, and it changed it. The debate has to go on. It's not climategate that will decide the future of climate science or climate policies. It's the process. Sounds frustrating, maybe.

Werner Krauss said...

@ Hans #43

"Well, that's how it is" - the authors quote several others like Traweek, Latour / Wolgar or Knorr-Cetina who show in ethnographic detail how informal and semi-informal communication influence the construction of facts in science. To have this documented and "frozen" in form of emails is really new and I think we have no experience how to deal with this.

In case you think that fraud is involved in climategate, of course, I would also judge the case as unusual. The authors are not really interested in this question, they only use the emails to document how facts are constructed.

Why not "climate science IN the anthropocene"? It's a geological era, and we live in that era. True, we still act as we could easily separate between nature and culture; between science and life; between laboratory and world and so on. In the anthropocene, this is by definition no longer possible. The earth is the laboratory, life is the experiment. Separations become muddy, human traces everywhere. No laboratory planet in sight to run tests. We have to decide on uncertain basis, even though both science and society pretend that we still live in the 19th century and rational science could solve all problems. That's what I call a challenge!

The quadruple "C" is not necessary. Because ALL facts are culturally constructed, as science studies show convincingly. Thus, "cultural" doesn't signify a difference any more. There are only good and bad constructions. Good and bad constructed hockey sticks. But the hockey stick won't decide the climate problem - which is frustrating for both sides. In the end, there is only one CC (formerly know as CCCC) as a huge challenge for both science and society, in terms of epistemology and ontology.

MikeR said...

@Werner - but please understand. It would never occur to me that this has anything to do with "the truth about climate science". That has to be left up to you guys.
What it does mean to me is what I heard Richard Muller say: There is a group of scientists I can't trust. I'd add, there's a larger group of scientists I have trouble trusting as well: the ones who keep defending the first group. I wish they'd stop.

Anonymous said...

Werner #9 "there is one special thing about climategate which indeed is a matter of concern: we don't know whether the thing with carbon based anthropogenic climate change is made-up or not."

That might be the case if the whole basis of anthropogenic climate change rested on the work of the emailers. But it doesn't. It rests on the work of thousands of scientist in multiple disciplines:

Atmospheric and Physical Sciences: Climatology, Meteorology, Atmospheric dynamics, Atmospheric physics, Atmospheric chemistry, Solar physics, Historical climatology

Earth Sciences: Geophysics, Geochemistry, Geology, Soil Science, Oceanography, Glaciology, Palaeoclimatology, Palaeoenvironmental reconstruction

Biological Sciences: Ecology, Synthetic biology, Biochemistry, Global change biology, Biogeography, Ecophysiology, Ecological genetics

Mathematics, Statistics and Computational analysis: Applied mathematics, Mathematical modelling, Computer science, Numerical modelling, Bayesian inference, Mathematical statistics, Time series analysis

etc, etc.... Are you saying that they're ALL making it up? Noooo... don't go there!

Werner Krauss said...

@ Anonymous #47

Thanks, you are perfectly right; I used sloppy language. Of course, those disciplines listed in your comment do not make it all up! They just do much more except focusing solely on carbon, and I guess, many of them would confirm that there is more than carbon to climate change.

What I had in mind was a quote from Jerry Ravetz, which goes like that:

“Are we really experiencing Anthropogenic Carbon-based Global Warming? If the public loses faith in that claim, then the situation of science will be altered for worse. There is very unlikely to be a crucial experience that either confirms of refutes the claim; the post-normal situation is just too complex. The consensus is likely to depend on how much trust can still be put in science” (Ravetz 2010).

Other greenhouse gases and factors also play a considerable role. The focus on carbon alone led to a kind of "war on carbon" - similar to the "war on terror", and Ravetz sees a danger in this, because specific politics (Kyoto, world summits, global treaties) and climate discourses (Al Gore's Inconvenient truth) rest mostly on the carbon-assumption. And there are limits to this carbon-based approach, as we could see at the limited success of Kyoto or the frustrating results of recent climate summits.

This is something totally different from "denying" the role of carbon; it is just pointing out a) that climate change is not identical with carbon emission only and b) the inherent dangers of a linear relation between science and climate politics solely based on carbon - which, in fact, was the case in the first part of the 21st century.

Thanks for the great list of those disciplines which contribute to climate science. Very useful indeed, I will keep it!

Maybe it is telling that you do not mention any social sciences, who maybe do not research the "naked" climate, but who considerably help to shape and adjust climate discourse and politics. There is more to climate than greenhouse gases only; or, to put it more poetically, those gases described by scientists also heat up the political greenhouse we inhabit - and what worth is science when it reduces climate change to carbon and thus maybe helps to support dilettantish politics?

In the context of this thread: climate change is a construction, as is the role of carbon. We have to be careful and should be aware that the truth is not "out there" in the atmosphere, but also in here, in the politics of science.

hvw said...

Werner, 9

"we don't know whether the thing with carbon based anthropogenic climate change is made-up or not. ... we don't know whether this is construction which will hold the test of time or not."

The theory of "carbon based anthropogenic climate change" is essentially unchanged since 1896.

Please tell, at which point do you consider a scientific construction of fact having passed the test of time and reached maturity such that it should be regarded as fact and not discussed anymore in the arena of decision making and politics?

Hans von Storch said...

hvw/49. Your claim "The theory of "carbon based anthropogenic climate change" is essentially unchanged since 1896." is a case of a broad claim, which depends very much on what you precisely mean. Do not forget that Arrhenius dealt with the anthropogenic climate change aspect only on the side, dismissing it as a very slow process (hundreds of years) because of a misunderstanding about the update of CO2 by the ocean.

Is the sensitivity of the system part of the theory?

Werner Krauss said...

@ hvn, 49

Don't feel too comfortable defending my statement again. Thin ice, and it's spring and getting warmer...

Obviously, version #48 didn't help to make my point more clear?

I'll give it another try. Carbon contributes to global warming - no doubt, box closed, we can go on.

But: Carbon is not the only factor that contributes to global warming: which other factors are relevant? How relevant?
And is the focus on carbon the best way to fight anthropogenic climate change? Or should we keep an eye on methane, but also on demographics, land use, energy consumption, unequal power relations? "Carbon based anthropogenic climate change": box far from closed, I'd say.

There is no general answer to the second part of your question. I would suggest (with Ludwik Fleck in mind), that "carbon based anthropogenic climate change" is a "true in a certain, limited perspective", a partial truth. True in a specific perspective, but not the truth about anthropogenic climate change.

Correct?

Hans von Storch said...

Anonymous/47.

Your list is impressive but misleading. Only a few of these groups are actually dealing with the issue if we have anthropogenic global warming going on (my group is among them), the rest is operating on the belief in this finding. It was a major problem with WG II of IPCC that they did not perform robust detection and attribution analysis - with the bottom line, that most of the scientists from your list are either believing what the D&A Group (IDAG, of which we are part) is telling them, or that they simply believe what is written by all kind of media.

I am personally convinced, by being involved in the statistical D&A work, that we see anthropogenic (GHG-related) climate change emerging in these decades. we can explain this change only by considering elevated GHG concentrations as a key driver of this change. As a decent scientist I add the caveat that this is my best present knowledge, and that there is always a (minimal) probability of error.

hvw said...

Hans von Storch, 50

More precisely, I meant a statement such as "Anthropogenic CO2 emisions have a global temperature effect, which is large and fast compared to natural processes and therefore potentially poses a serious hazard to human societies globally".

Arrhenius did not foresee the magnitude of the anthropogenic effect, as far as I can tell, because he did not foresee the dramatic increase of emissions. His sensitivity estimates are a bit larger than what we assume today.

Curious: Where exactly did you get the "misunderstanding of ocean uptake" from?

To answer you question: I include the order of magnitude of the sensitivity, which allows us to judge whether we have a problem or not, in "the constructed fact" that hasn't changed since > 100 years.

hvw said...

Werner Krauss, 51

"Obviously, version #48 didn't help to make my point more clear?'

It would have, had I read it before posting ...

I think the Ravetz quote comes from a very special postnormal context. Your using it in a climategate thread puts you, for the casual reader, square into the "What can a 0.005% concentration of CO2 possibly do?"-camp, where you don't want to be.

I don't know what is known about the relative effect of other GHGs, currently, but it seems clear that any effective mitigation strategy's big problem is CO2 in any case.

For the rest, I completely agree with you. A tunnel-vision on carbon for finding solutions and strategies, and the atmospheric chemistry dominating the politics-science interface has proven to be not helpful and there is the need for re-framing to think better, etc... Thats partly the raison d'être for this blog I guess.